A tea processing quality control method and system based on image processing

Abstract

The present invention is applicable to the fields of tea processing and computer technology, and provides a tea processing quality control method and system based on image processing, the method comprising the following steps: acquiring the status information of the current tea in real time; the status information comprising an appearance image, glossiness and water content; determining the chromaticity of the current tea according to the appearance image of the current tea; determining the relative chromaticity value of the current tea according to the chromaticity of the current tea based on a preset standard tea chromaticity; obtaining the status value of the current tea according to the glossiness, water content and relative chromaticity value of the current tea; predicting the quality value of the current tea after the tea is withered for a certain period of time at the current withering temperature in real time according to the status value of the current tea based on a preset quality value prediction model; and adjusting the withering temperature and / or withering time in real time according to the predicted quality value.

Description

Technical Field

[0001] The invention belongs to the field of tea processing and computer technology, and in particular relates to a tea processing quality control method and system based on image processing. Background Art

[0002] When tea is processed, it is generally necessary to go through the process of killing green tea leaves. The main purpose of killing green tea leaves is to destroy and passivate the oxidase activity in fresh tea leaves through high temperature, and reduce the water content of tea leaves, so as to inhibit the enzymatic oxidation of tea polyphenols and other substances in tea leaves, promote the formation of good aroma, and soften the tea leaves for subsequent processing. By controlling the degree of killing green tea leaves during processing, the quality of tea products can be well controlled.

[0003] In the existing tea withering process, the process parameters such as the temperature and time of withering are set according to the type of tea, and are generally kept unchanged, or adjusted through manual observation; however, in the actual processing process, under the same process conditions, the current state of the tea will have a certain impact on the degree of withering. If the withering process remains unchanged, it is difficult to control the degree of withering of the tea, and thus the quality of the final tea product cannot be guaranteed; in addition, the traditional method of manually observing the state of the tea in real time, although the withering process can be adjusted accordingly, it has problems such as low efficiency and low accuracy, so it is still in urgent need of improvement. Summary of the invention

[0004] The purpose of the embodiment of the present invention is to provide a tea processing quality control method based on image processing, aiming to solve the above technical problems.

[0005] The embodiment of the present invention is implemented as follows: a tea processing quality control method based on image processing, which comprises the following steps:

[0006] Acquire the status information of the current tea leaves in real time; the status information includes appearance image, glossiness and water content;

[0007] Determine the chromaticity of the current tea leaves according to the appearance image of the current tea leaves;

[0008] Based on the preset standard tea color, determine the relative color value of the current tea according to the color of the current tea;

[0009] According to the glossiness, water content and relative chromaticity value of the current tea leaves, the state value of the current tea leaves is obtained;

[0010] Based on the preset quality value prediction model, the quality value of the current tea leaves after being withered for a certain period of time at the current withering temperature is predicted in real time according to the current state value of the tea leaves;

[0011] Adjust the fixing temperature and / or fixing time in real time according to the predicted quality value.

[0012] Preferably, the step of determining the chromaticity of the current tea leaves according to the appearance image of the current tea leaves specifically includes:

[0013] Select several pixels from the current tea image;

[0014] Determine the RGB values ​​of a plurality of pixels respectively; the RGB values ​​include R value, G value and B value;

[0015] The chromaticity of the current tea is determined based on the R value, G value and B value of several pixel points.

[0016] Preferably, the step of determining the chromaticity of the current tea leaves according to the R value, G value and B value of a plurality of pixel points specifically comprises:

[0017] According to the R values ​​of several pixel points, the average R value of the current tea is calculated;

[0018] According to the G values ​​of several pixel points, the average G value of the current tea is calculated;

[0019] According to the B values ​​of several pixel points, the average B value of the current tea is calculated;

[0020] The chromaticity of the current tea is determined based on the average R value, average G value and average B value of the current tea.

[0021] Preferably, the standard tea chromaticity includes the R value, the G value and the S value of the standard tea; the calculation formula of the relative chromaticity value of the current tea is:

[0022] ;

[0023] Where c0 is the relative chromaticity value of the current tea; n is the total number of pixels obtained; R i is the R value of the i-th pixel in the current tea appearance image, G i is the G value of the i-th pixel in the current tea appearance image, B i is the B value of the i-th pixel in the current tea appearance image; R c is the R value of the preset standard tea; G c is the G value of the preset standard tea; B c It is the B value of the preset standard tea.

[0024] Preferably, the method for constructing the quality value prediction model comprises the following steps:

[0025] Acquire the initial state values ​​of several groups of historical records of tea leaves and their quality values ​​after being treated with withering at a certain temperature for a certain period of time; the quality value is determined according to the evaluation value of the tea leaves after withering; the evaluation value includes a relative state value and an aroma value; the relative state value is determined according to the state values ​​of the tea leaves before and after withering; the aroma value is determined according to the concentration of aromatic substances in the tea leaves after withering;

[0026] According to several groups of historical records of the initial state values ​​of tea leaves and the quality values ​​after being treated at a certain temperature for a certain period of time, a functional relationship between the quality value of tea leaves and the state value, the temperature and the time of tea leaves is established;

[0027] Based on the Kriging interpolation method, a quality value prediction model was established according to the functional relationship between the quality value of tea leaves and the state value, withering temperature and withering time.

[0028] Preferably, the quality value prediction model is:

[0029] ;

[0030] Where Q0' refers to the predicted value of the quality of tea leaves with a current state value of s0 after being treated at the current withering temperature T0 for a certain time t0; Q i It means that the i-th historically known state value is s i The tea leaves are fixed at the temperature T i Perform the fixing treatment for a certain time t i The quality value after is 1≤i≤m; is the Kriging coefficient.

[0031] Preferably, the Kriging coefficient satisfies that the variance between the predicted value and the true value of the quality value is minimal, and the mean of the difference between the predicted value and the true value of the quality value is zero; the method for determining the Kriging coefficient comprises the following steps:

[0032] Based on the Lagrange multiplier method, the solution matrix of the Kriging coefficient is obtained according to the covariance and semivariance between the historically known quality values;

[0033] Based on the historically known state values, killing temperature and killing time, define the correlation between the historically known quality values;

[0034] According to the correlation and semivariance between the quality values ​​known in history, the functional relationship between the semivariance and the correlation is fitted;

[0035] According to the current state value, withering temperature and withering time of tea leaves, define the correlation between the quality value to be predicted and the known quality value;

[0036] Based on the functional relationship between the semivariance and the correlation, according to the correlation between the quality value to be predicted and the known quality value, the semivariance estimation between the quality value to be predicted and the known quality value is obtained;

[0037] Based on the solution matrix of the Kriging coefficient, the Kriging coefficient is determined according to the semi-variance between the historically known quality values ​​and the semi-variance estimate between the quality value to be predicted and the known quality value.

[0038] Another object of the embodiment of the present invention is to provide a tea processing quality control system based on image processing, which is used to implement the above-mentioned tea processing quality control method based on image processing, and comprises:

[0039] A status information acquisition unit, used to acquire the status information of the current tea leaves in real time; the status information includes appearance image, glossiness and water content;

[0040] A chromaticity determination unit, used to determine the chromaticity of the current tea leaves according to the appearance image of the current tea leaves;

[0041] A relative chromaticity value calculation unit, used to determine the relative chromaticity value of the current tea leaves according to the chromaticity of the current tea leaves based on the preset standard tea chromaticity;

[0042] A state value calculation unit, used to obtain the state value of the current tea leaves according to the glossiness, water content and relative chromaticity value of the current tea leaves;

[0043] A quality value prediction unit, for predicting in real time the quality value of the current tea leaves after being withered at the current withering temperature for a certain period of time according to the state value of the current tea leaves based on a preset quality value prediction model;

[0044] The process adjustment unit is used to adjust the fixing temperature and / or fixing time in real time according to the predicted quality value.

[0045] Preferably, the chromaticity determination unit comprises:

[0046] A pixel point selection module is used to select a number of pixel points from the current tea image;

[0047] The RGB value determination module is used to respectively determine the RGB values ​​of a plurality of pixel points; the RGB values ​​include R value, G value and B value;

[0048] The chromaticity calculation module is used to determine the chromaticity of the current tea according to the R value, G value and B value of a number of pixel points.

[0049] Preferably, the tea processing quality control system also includes:

[0050] A historical data acquisition unit, used to acquire the initial state values ​​of tea leaves of several groups of historical records and the quality values ​​of tea leaves after being treated with tea leaves at a certain temperature for a certain period of time;

[0051] A function relationship establishing unit, for establishing a function relationship between the quality value of the tea leaves and the state value, the withering temperature and the withering time according to a plurality of groups of historical records of the initial state value of the tea leaves and the quality value after withering at a certain withering temperature for a certain period of time;

[0052] The model building unit is used to establish a quality value prediction model based on the Kriging interpolation method according to the functional relationship between the quality value of tea leaves and the state value, the withering temperature and the withering time.

[0053] An image processing-based quality control method for tea processing provided in an embodiment of the present invention establishes a quality value prediction model and monitors and obtains status information such as the image, glossiness and water content of the current tea in real time. The quality value of the current tea after withering at the current withering temperature for a certain period of time can be predicted in real time through these status information. Then, the withering process conditions such as the withering temperature and withering time can be adjusted in real time according to the predicted quality value to ensure the quality of the tea after withering. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 A schematic diagram of a process flow of a tea processing quality control method based on image processing provided by an embodiment of the present invention;

[0055] Figure 2 A schematic diagram of a specific flow chart of step S200 provided in an embodiment of the present invention;

[0056] Figure 3 A schematic diagram of a specific flow chart of step S230 provided in an embodiment of the present invention;

[0057] Figure 4 A schematic diagram of a process for constructing a quality value prediction model provided by an embodiment of the present invention;

[0058] Figure 5 A schematic diagram of a flow chart of a method for determining a Kriging coefficient provided in an embodiment of the present invention;

[0059] Figure 6 A schematic diagram of a tea processing quality control system based on image processing provided by an embodiment of the present invention;

[0060] Figure 7 A schematic diagram of a chromaticity determination unit provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0061] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0062] like Figure 1 As shown, in one embodiment of the present invention, a tea processing quality control method based on image processing is provided, which comprises the following steps:

[0063] S100, obtaining status information of the current tea leaves in real time; the status information includes appearance image, glossiness and water content;

[0064] S200, determining the chromaticity of the current tea leaves according to the appearance image of the current tea leaves;

[0065] S300, determining the relative chromaticity value of the current tea leaves according to the chromaticity of the current tea leaves based on the preset standard tea chromaticity;

[0066] S400, obtaining a state value of the current tea leaves according to the glossiness, water content and relative chromaticity value of the current tea leaves;

[0067] S500, based on a preset quality value prediction model, predicting in real time the quality value of the current tea leaves after being withered at the current withering temperature for a certain period of time according to the state value of the current tea leaves;

[0068] S600: adjusting the fixing temperature and / or fixing time in real time according to the predicted quality value.

[0069] In an embodiment of the present invention, by establishing a quality value prediction model and real-time monitoring and acquiring status information such as the image, glossiness and moisture content of the current tea leaves, the quality value of the current tea leaves after being withered at the current withering temperature for a certain period of time can be predicted in real time through these status information. Then, the withering process conditions such as the withering temperature and withering time can be adjusted in real time according to the predicted quality value to ensure the quality of the tea leaves after withering.

[0070] In step S100, the current status information of the tea leaves refers to the status information of the tea leaves undergoing the withering process; it should be noted that the method provided in the embodiment of the present invention does not limit the method of withering the tea leaves, that is, it is applicable to withering processes such as pan withering, drum withering, and microwave withering, but is not limited to this; in practical applications, the appearance image of the tea leaves undergoing the withering process can be obtained by an industrial camera; the glossiness can be determined by testing the reflective ability of the tea surface with an existing gloss meter; the moisture content can be determined by a drying method or an infrared moisture meter.

[0071] As a preferred embodiment of the present invention, Figure 2As shown, the step of determining the chromaticity of the current tea leaves (i.e., step S200) according to the appearance image of the current tea leaves specifically includes:

[0072] S210, selecting a number of pixel points from the current tea image;

[0073] S220, respectively measuring RGB values ​​of a plurality of pixel points; the RGB values ​​include R value, G value and B value;

[0074] S230, determining the chromaticity of the current tea leaves according to the R value, G value and B value of a plurality of pixel points.

[0075] Among them, ‌RGB value‌ refers to the numerical value used to represent different colors in the RGB color mode; the ‌RGB value of each color is jointly determined by the intensity values ​​of the three color channels of red (R), green (G), and blue (B); the intensity values ​​of the three color channels of R, G, and B range from 0 to 255, where 255 represents the maximum intensity and 0 represents the minimum intensity. In the embodiment of the present invention, since the appearance of the tea leaves before withering is generally light green, its G value is relatively large; and the appearance of the tea leaves after withering is generally reddish or yellowish dark green, its G value is generally reduced, and its R value is generally increased.

[0076] As a preferred embodiment of the present invention, Figure 3 As shown, the step of determining the chromaticity of the current tea leaves (i.e., step S230) according to the R value, G value, and B value of a plurality of pixel points specifically includes:

[0077] S231, calculating the average R value of the current tea according to the R values ​​of a plurality of pixel points;

[0078] S232, calculating the average G value of the current tea according to the G values ​​of a plurality of pixel points;

[0079] S233, calculating the average B value of the current tea according to the B values ​​of a plurality of pixel points;

[0080] S234. Determine the chromaticity of the current tea leaves according to the average R value, average G value and average B value of the current tea leaves.

[0081] Specifically, the average RGB value of the current tea is used as the chromaticity of the current tea, and the chromaticity of the current tea can be recorded as:

[0082] ;

[0083] Where R0 is the average R value of the current tea leaves; G0 is the average G value of the current tea leaves; B0 is the average B value of the current tea leaves; n is the total number of pixels obtained; R i is the R value of the i-th pixel in the current tea appearance image, G iis the G value of the i-th pixel in the current tea appearance image, B i is the B value of the i-th pixel in the current tea appearance image.

[0084] As a preferred embodiment of the present invention, in step S300, the standard tea chromaticity includes the R value, the G value and the S value of the standard tea; the calculation formula of the relative chromaticity value of the current tea is:

[0085] ;

[0086] Where c0 is the relative chromaticity value of the current tea; R c is the R value of the preset standard tea; G c is the G value of the preset standard tea; B c It is the B value of the preset standard tea.

[0087] It should be noted that the R value, G value and B value of the above-mentioned standard tea refer to the R value, G value and B value of qualified tea after withering, which can generally be preset according to different types of tea.

[0088] As a preferred embodiment of the present invention, in step S400, the calculation formula of the current tea state value is as follows:

[0089] ;

[0090] Wherein, s0 is the state value of the current tea leaves, g0 is the glossiness of the current tea leaves, and w0 is the water content of the current tea leaves; k1, k2 and k3 are weight coefficients of relative chromaticity value, glossiness and water content, respectively, which can be determined according to the type of tea leaves. Generally, k1=0.4-0.6, k2=0.2-0.3, k3=0.2-0.3, and the sum of the three is 1, but not limited to this;

[0091] As a preferred embodiment of the present invention, in step S500, Figure 4 As shown, the method for constructing the quality value prediction model includes the following steps:

[0092] S510, obtaining the initial state values ​​of several groups of historical records of tea leaves and their quality values ​​after being treated with withering at a certain temperature for a certain period of time; the quality value is determined according to the evaluation value of the tea leaves after withering; the evaluation value includes a relative state value and an aroma value; the relative state value is determined according to the state values ​​of the tea leaves before and after withering; the aroma value is determined according to the concentration of aromatic substances in the tea leaves after withering;

[0093] S520, establishing a tea quality value Q according to the initial state values ​​of the tea leaves recorded in several groups of history and the quality values ​​after the tea leaves are treated at a certain temperature for a certain period of time. i About the state value s i , Fixing temperature T i and finishing time t i The functional relationship is as follows:

[0094] ;

[0095] S530. Based on the Kriging interpolation method, a quality value prediction model is established according to the functional relationship between the quality value of the tea leaves and the state value, the withering temperature and the withering time.

[0096] In practical applications, the quality value of the tea leaves after withering can be obtained by weighted summation based on the relative state value and aroma value of the tea leaves after withering, as follows:

[0097] ;

[0098] In the formula, Q i It means that the i-th historically known state value is s i The tea leaves are fixed at the temperature T i Perform the fixing treatment for a certain time t i The quality value after is the state value of the tea leaves after withering, and its calculation method is the same as the calculation method of the state value of the current tea leaves mentioned above; It is the standard state value of the tea leaves after withering, that is, the state value of qualified tea leaves after withering, which can be preset according to different types of tea leaves; a i is the aroma value of the tea leaves after withering, which can be determined according to the concentration of aromatic substances in the tea leaves after withering; specifically, the aroma value of the tea leaves after withering a i It is the ratio of the concentration of aromatic substances in the tea leaves after withering to the preset standard concentration (the preset standard concentration refers to the concentration of aromatic substances in qualified tea leaves, which can be preset according to different types of tea leaves); k4 and k5 are weight coefficients, which can be determined according to actual conditions.

[0099] As a preferred embodiment of the present invention, the above-mentioned quality value prediction model is:

[0100] ;

[0101] Where Q0' refers to the predicted value (estimate) of the quality value of tea leaves with a current state value of s0 after being treated at the current withering temperature T0 for a certain time t0; Q i It means that the i-th historically known state value is s i The tea leaves are fixed at the temperature T iPerform the fixing treatment for a certain time t i The quality value after is 1≤i≤m; is the Kriging coefficient.

[0102] As a preferred embodiment of the present invention, the Kriging coefficient satisfies that the variance of the predicted value Q0' and the true value Q0 of the quality value is minimal, and the mean of the difference between the predicted value Q0' and the true value Q0 of the quality value is zero; Figure 5 As shown, the method for determining the Kriging coefficient includes the following steps:

[0103] S531, based on the Lagrange multiplier method, according to the covariance and semivariance between the historically known quality values, a solution matrix of the Kriging coefficient is obtained;

[0104] S532, defining the correlation between the historically known quality values ​​according to the historically known state values, the fixing temperature and the fixing time;

[0105] S533, according to the correlation and semi-variance between the quality values ​​known in history, fitting a functional relationship between the semi-variance and the correlation;

[0106] S534, defining the correlation between the quality value to be predicted and the known quality value according to the current state value of the tea leaves, the withering temperature and the withering time;

[0107] S535, based on the functional relationship between the semivariance and the correlation degree, according to the correlation degree between the quality value to be predicted and the known quality value, obtain the semivariance estimation between the quality value to be predicted and the known quality value;

[0108] S536. Based on the solution matrix of the Kriging coefficients, the Kriging coefficients are determined according to the semi-variance between the historically known quality values ​​and the semi-variance estimates between the quality values ​​to be predicted and the known quality values.

[0109] Specifically, the Kriging coefficient The constraints are as follows: Kriging coefficient To satisfy the optimal set of coefficients that minimize the variance Var(Q0'-Q0) between the predicted value Q0' and the true value Q0, and at the same time satisfy the condition of unbiased estimation, that is, the mean value E(Q0'-Q0) of the difference between the predicted value Q0' and the true value Q0 = 0; based on this, .

[0110] Then, define the known quality value Q obtained historically i and Q j The semivariogram function between is as follows:

[0111] ;

[0112] in, Refers to the state value s i The tea leaves are fixed at the temperature T i The fixing temperature is t i The quality value after Refers to the state value s j The tea leaves are fixed at the temperature T j The fixing temperature is t j The quality value after The known quality value Q obtained from history i and Q j The covariance between The known quality value Q obtained from history i and Q j The variance between .

[0113] It should be noted that since any two known quality values ​​Q i and Q j The difference between them is a constant, so for any two known quality values ​​Q i and Q j The semivariance between satisfies:

[0114] .

[0115] Based on the Lagrange multiplier method, under the above constraints, Var(Q0'-Q0) is minimized, and the following solution matrix of the Kriging coefficient can be obtained:

[0116] ;

[0117] Inverting the above matrix can obtain the Kriging coefficient A set of optimal solutions.

[0118] in, is the Lagrange multiplier; r i0 Q i and the semivariance between Q0; r i0 The calculation method is as follows:

[0119] First, define Q i and Q j The correlation degree d ij for:

[0120] ;

[0121] Based on historical data, As the vertical axis, As the horizontal axis, draw a scatter plot to find the best fitting curve and The functional relationship between semivariance and correlation is obtained: ;

[0122] Next, define Q0 and Q i The correlation between i0 for:

[0123] ;

[0124] According to the above functional relationship, substitute , we can get r i0 .

[0125] As a preferred embodiment of the present invention, in step S600, the predicted quality value Q0' and the preset standard tea quality value Q c (Generally, Q c =1), and find the semivariance r of the two c0 :

[0126] ;

[0127] At the same time, define the quality values ​​Q0 and Q c The correlation between c0 :

[0128] ;

[0129] Among them, s c =s0,T c is the adjusted fixing temperature, t c is the adjusted time, that is, (T c -T0) is the adjustment range of the fixing temperature, (t c -t0) is the adjustment range of the finishing time; according to the above functional relationship between semivariance and correlation, substitute r c0 , we can inversely calculate d c0 , then according to the above correlation formula, we can get (T c -T0) and (t c -t0), and the fixing process is adjusted in real time accordingly; it should be noted that (T c -T0) and (t c -t0) can be positive or negative depending on the actual situation and the current state of the tea leaves Determine, as long as the predicted quality value Q0'≥Q under the adjusted killing process conditions is met c That's it.

[0130] like Figure 6As shown, in another embodiment of the present invention, a tea processing quality control system based on image processing is also provided, which is used to implement the above-mentioned tea processing quality control method based on image processing, and includes:

[0131] The status information acquisition unit 10 is used to acquire the status information of the current tea leaves in real time; the status information includes appearance image, glossiness and water content;

[0132] A chromaticity determination unit 20, for determining the chromaticity of the current tea leaves according to the appearance image of the current tea leaves;

[0133] The relative chromaticity value calculation unit 30 is used to determine the relative chromaticity value of the current tea leaves according to the chromaticity of the current tea leaves based on the preset standard tea chromaticity;

[0134] The state value calculation unit 40 is used to obtain the state value of the current tea leaves according to the glossiness, water content and relative chromaticity value of the current tea leaves;

[0135] The quality value prediction unit 50 is used to predict the quality value of the current tea leaves after being withered for a certain period of time at the current withering temperature according to the state value of the current tea leaves in real time based on a preset quality value prediction model;

[0136] The process adjustment unit 60 is used to adjust the fixing temperature and / or fixing time in real time according to the predicted quality value.

[0137] As a preferred embodiment of the present invention, Figure 7 As shown, the chromaticity determination unit 20 includes:

[0138] A pixel point selection module 21 is used to select a number of pixel points from the current tea image;

[0139] RGB value determination module 22, used to respectively determine the RGB values ​​of a plurality of pixel points; the RGB values ​​include R value, G value and B value;

[0140] The chromaticity calculation module 23 is used to determine the chromaticity of the current tea according to the R value, G value and B value of a plurality of pixel points.

[0141] In another preferred embodiment of the present invention, the tea processing quality control system further comprises:

[0142] A historical data acquisition unit, used to acquire the initial state values ​​of tea leaves of several groups of historical records and the quality values ​​of tea leaves after being withered at a certain withering temperature for a certain period of time;

[0143] A function relationship establishing unit, for establishing a function relationship between the quality value of the tea leaves and the state value, the withering temperature and the withering time according to a plurality of groups of historical records of the initial state value of the tea leaves and the quality value after withering at a certain withering temperature for a certain period of time;

[0144] The model building unit is used to establish a quality value prediction model based on the Kriging interpolation method according to the functional relationship between the quality value of tea leaves and the state value, the withering temperature and the withering time.

[0145] It should be noted that each of the above-mentioned units can be implemented in the form of a computer program, which can be run on a computer device. The memory of the computer device can store the computer program constituting each module so that the processor executes each step of the above-mentioned tea processing quality control method based on image processing.

[0146] It should be understood that, although each step in the flow chart of each embodiment of the present invention is shown in sequence according to the indication of the arrow, these steps are not necessarily performed in sequence according to the order indicated by the arrow. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be performed in other orders. Moreover, at least a portion of the steps in each embodiment may include a plurality of sub-steps or a plurality of stages, and these sub-steps or stages are not necessarily performed at the same time, but can be performed at different times, and the execution order of these sub-steps or stages is not necessarily performed in sequence, but can be performed in turn or alternately with at least a portion of other steps or sub-steps or stages of other steps.

[0147] A person skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program, and the program can be stored in a non-volatile computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory.

[0148] The above-mentioned embodiments only express several implementation methods of the present invention, and the description thereof is relatively specific and detailed, but it cannot be understood as limiting the scope of the patent of the present invention. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.

Claims

1. A tea processing quality control method based on image processing, characterized in that: The following steps are involved: Acquire the current status information of tea leaves in real time; the status information includes appearance image, glossiness and water content; Determine the chromaticity of the current tea leaves according to the appearance image of the current tea leaves; Based on the preset standard tea color, determine the relative color value of the current tea according to the color of the current tea; According to the glossiness, water content and relative chromaticity value of the current tea leaves, the state value of the current tea leaves is obtained; Based on the preset quality value prediction model, the quality value of the current tea leaves after being withered for a certain period of time at the current withering temperature is predicted in real time according to the current state value of the tea leaves; Adjust the fixing temperature and / or fixing time in real time according to the predicted quality value; The step of determining the chromaticity of the current tea leaves according to the appearance image of the current tea leaves specifically includes: Select several pixels from the current tea image; Determine the RGB values ​​of a plurality of pixels respectively; the RGB values ​​include R value, G value and B value; Determine the chromaticity of the current tea leaves based on the R value, G value and B value of a number of pixel points; The steps of determining the chromaticity of the current tea leaves according to the R value, G value and B value of a plurality of pixel points specifically include: According to the R values ​​of several pixel points, the average R value of the current tea is calculated; According to the G values ​​of several pixel points, the average G value of the current tea is calculated; According to the B values ​​of several pixel points, the average B value of the current tea is calculated; Determine the chromaticity of the current tea leaves according to the average R value, average G value and average B value of the current tea leaves; The standard tea chromaticity includes the R value, G value and B value of the standard tea; the calculation formula of the relative chromaticity value of the current tea is: Where c0 is the relative chromaticity value of the current tea; n is the total number of pixels obtained; R i is the R value of the i-th pixel in the current tea appearance image, G i is the G value of the i-th pixel in the current tea appearance image, B i is the B value of the i-th pixel in the current tea appearance image; R c is the R value of the preset standard tea; G c is the G value of the preset standard tea; B c is the B value of the preset standard tea; The calculation formula of the current tea state value is as follows: s0 = k1c0 + k2g0 + k3w0; Where s0 is the current state value of the tea leaves, g0 is the glossiness of the tea leaves, and w0 is the water content of the tea leaves; k1, k2, and k3 are the weight coefficients of the relative chromaticity value, glossiness, and water content, respectively, and the sum of the three is 1; The method for constructing the quality value prediction model comprises the following steps: Obtaining several groups of historical records of initial state values ​​of tea leaves and quality values ​​after being treated with withering at a certain temperature for a certain period of time; the quality value is determined according to the evaluation value of the tea leaves after withering; the evaluation value includes a relative state value and an aroma value; the relative state value is determined according to the state values ​​of the tea leaves before and after withering; the aroma value is determined according to the concentration of aromatic substances in the tea leaves after withering; According to several groups of historical records of the initial state values ​​of tea leaves and their quality values ​​after being treated at a certain temperature for a certain period of time, a functional relationship between the quality value of tea leaves and the state value, the temperature and the time of tea leaves is established; among which the i-th historically known state value is s i The tea leaves are fixed at the temperature T i Perform the fixing treatment for a certain time t i The quality value Q i The calculation method is as follows: In the formula, s i ′ is the state value of the tea leaves after withering, and its calculation method is the same as the calculation method of the state value of the current tea leaves mentioned above; s c ′ is the standard state value of tea leaves after withering; a i is the aroma value of the tea leaves after withering, specifically the ratio of the concentration of aromatic substances in the tea leaves after withering to the preset standard concentration; k4 and k5 are weight coefficients; Based on the Kriging interpolation method, a quality value prediction model is established according to the functional relationship between the quality value of tea leaves and the state value, the withering temperature and the withering time. The quality value prediction model is: Where Q0' refers to the predicted value of the quality of tea leaves with a current state value of s0 after being treated at the current withering temperature T0 for a certain time t0; Q i It means that the i-th historically known state value is s i The tea leaves are fixed at the temperature T i Perform the fixing treatment for a certain time t i The quality value after, 1≤i≤m; λ i is the Kriging coefficient.

2. The tea processing quality control method based on image processing according to claim 1, characterized in that: The Kriging coefficient satisfies that the variance between the predicted value and the true value of the quality value is minimal, and at the same time satisfies that the mean of the difference between the predicted value and the true value of the quality value is zero; the method for determining the Kriging coefficient comprises the following steps: Based on the Lagrange multiplier method, the solution matrix of the Kriging coefficient is obtained according to the covariance and semivariance between the historically known quality values; Based on the historically known state values, killing temperature and killing time, define the correlation between the historically known quality values; According to the correlation and semivariance between the quality values ​​known in history, the functional relationship between the semivariance and the correlation is fitted; According to the current state value, withering temperature and withering time of tea leaves, define the correlation between the quality value to be predicted and the known quality value; Based on the functional relationship between the semivariance and the correlation, according to the correlation between the quality value to be predicted and the known quality value, the semivariance estimation between the quality value to be predicted and the known quality value is obtained; Based on the solution matrix of the Kriging coefficients, the Kriging coefficients are determined according to the semi-variance between the historically known quality values ​​and the semi-variance estimates between the quality values ​​to be predicted and the known quality values.

3. A tea processing quality control system based on image processing, used to implement the tea processing quality control method based on image processing according to any one of claims 1-2, characterized in that: include: A status information acquisition unit, used to acquire the status information of the current tea leaves in real time; The state information includes appearance image, glossiness and water content; A chromaticity determination unit, used to determine the chromaticity of the current tea leaves according to the appearance image of the current tea leaves; A relative chromaticity value calculation unit, used to determine the relative chromaticity value of the current tea leaves according to the chromaticity of the current tea leaves based on the preset standard tea chromaticity; A state value calculation unit, used to obtain the state value of the current tea leaves according to the glossiness, water content and relative chromaticity value of the current tea leaves; A quality value prediction unit, for predicting in real time the quality value of the current tea leaves after being withered at the current withering temperature for a certain period of time according to the state value of the current tea leaves based on a preset quality value prediction model; A process adjustment unit, used to adjust the fixing temperature and / or fixing time in real time according to the predicted quality value; A historical data acquisition unit, used to acquire several groups of historical records of initial state values ​​of tea leaves and quality values ​​after being treated with withering at a certain temperature for a certain period of time; the quality value is determined according to the evaluation value of the tea leaves after withering; the evaluation value includes a relative state value and an aroma value; the relative state value is determined according to the state value of the tea leaves before and after withering; the aroma value is determined according to the concentration of aromatic substances in the tea leaves after withering; The function relationship establishment unit is used to establish the function relationship between the quality value of the tea leaves and the state value, the withering temperature and the withering time according to the initial state value of the tea leaves recorded in several groups of historical records and the quality value after withering at a certain withering temperature for a certain time; wherein the i-th historically known state value is s i The tea leaves are fixed at the temperature T i Perform the fixing treatment for a certain time t i The quality value Q i The calculation method is as follows: In the formula, s i ′ is the state value of the tea leaves after withering, and its calculation method is the same as that of the current tea leaves; s c ′ is the standard state value of tea leaves after withering; a i is the aroma value of the tea leaves after withering, specifically the ratio of the concentration of aromatic substances in the tea leaves after withering to the preset standard concentration; k4 and k5 are weight coefficients; A model building unit, for building a quality value prediction model based on the Kriging interpolation method according to the functional relationship between the quality value of tea leaves and the state value, the withering temperature and the withering time; The quality value prediction model is: Where Q0' refers to the predicted value of the quality of tea leaves with a current state value of s0 after being treated at the current withering temperature T0 for a certain time t0; Q i It means that the i-th historically known state value is s i The tea leaves are fixed at the temperature T i Perform the fixing treatment for a certain time t i The quality value after, 1≤i≤m; λ i is the Kriging coefficient; The chromaticity determination unit comprises: A pixel point selection module is used to select a number of pixel points from the current tea image; The RGB value determination module is used to respectively determine the RGB values ​​of a plurality of pixel points; the RGB values ​​include R value, G value and B value; The chromaticity calculation module is used to determine the chromaticity of the current tea according to the R value, G value and B value of several pixel points; specifically used for: According to the R values ​​of several pixel points, the average R value of the current tea is calculated; According to the G values ​​of several pixel points, the average G value of the current tea is calculated; According to the B values ​​of several pixel points, the average B value of the current tea is calculated; Determine the chromaticity of the current tea leaves according to the average R value, average G value and average B value of the current tea leaves; The standard tea chromaticity includes the R value, G value and B value of the standard tea; the calculation formula of the relative chromaticity value of the current tea is: Where c0 is the relative chromaticity value of the current tea; n is the total number of pixels obtained; R i is the R value of the i-th pixel in the current tea appearance image, G i is the G value of the i-th pixel in the current tea appearance image, B i is the B value of the i-th pixel in the current tea appearance image; R c is the R value of the preset standard tea; G c is the G value of the preset standard tea; B c is the B value of the preset standard tea; The calculation formula of the current tea state value is as follows: s0 = k1c0 + k2g0 + k3w0; Where s0 is the current state value of the tea leaves, g0 is the glossiness of the tea leaves, and w0 is the water content of the tea leaves. k1, k2, and k3 are the weight coefficients of relative chromaticity value, glossiness, and water content, respectively, and their sum is 1.

Images